Comparative analysis and statistical performance evaluation of fractional PID controllers in fuel cell choppers optimized with metaheuristic algorithms

Fuel cells (FCs) are significant renewable energy sources that produce electrical energy, similar to photovoltaic solar panels and wind turbines. These cells use hydrogen as a fuel and produce pure water as waste, making them potential alternatives to batteries. This paper compares the parameters of a buck-type converter designed for use in fuel cell (FC) energy production systems. These parameters were optimized using five optimization algorithms. To ensure high quality, the converters that adjust the desired current and voltage values on the load must be controlled optimally. In this context, fractional-order PID (FOPID) controllers, which are advanced versions of standard PID controllers, were used to adjust the voltage of a load to its desired reference value. Five parameters (K_P, K_I, K_D, \(\mu\) $\mu$ , and \(\lambda\) $\lambda$ ) of this controller were adjusted using various algorithms: artificial bee colony (ABC) optimization, atom search optimization (ASO), Henry gas solubility optimization (HGSO), the JAYA algorithm, and the sine cosine algorithm (SCA) optimization. Among these algorithms, the HGSO algorithm demonstrated the best results. However, after a comprehensive statistical analysis, different insights emerged. The findings are then presented in a comparative analysis. Simulation studies were performed in the MATLAB/Simulink environment and the coding sections of MATLAB. The system under study comprises an FC, a converter, and loads. The responsiveness of the controller to variables and disturbances was examined by continuously adjusting both the reference value and load. The results derived from these calculations provide a detailed analysis of the controller’s ability to ensure that the voltage on the load aligns with the reference values.